Non-rigid 3D Shape Registration using an Adaptive Template

Hang Dai; Nicholas Edwin Pears; William Alfred Peter Smith

Non-rigid 3D Shape Registration using an Adaptive Template

Hang Dai, Nicholas Edwin Pears, William Alfred Peter Smith

Computer Science

Research output: Contribution to conference › Paper › peer-review

Abstract

We present a new fully-automatic non-rigid 3D shape registration (morphing) framework comprising (1) a new 3D landmarking and pose normalisation method; (2) an adaptive shape template method to improve the convergence of registration algorithms and achieve a better final shape correspondence and (3) a new iterative registration method that combines Iterative Closest Points with Coherent Point Drift (CPD) to achieve a more stable and accurate correspondence establishment than standard CPD. We call this new morphing approach \emph{Iterative Coherent Point Drift} (ICPD). Our proposed framework is evaluated qualitatively and quantitatively on three datasets: Headspace, BU3D and a synthetic LSFM dataset, and is compared with several other methods. The proposed framework is shown to give state-of-the-art performance.

Original language	English
Number of pages	15
Publication status	Published - 14 Sept 2018
Event	ECCV 2018: PeopleCap Workshop - GASTEIG Cultural Center, Munich, Germany Duration: 14 Sept 2018 → 14 Sept 2018 https://peoplecap2018.weebly.com/

Workshop

Workshop	ECCV 2018: PeopleCap Workshop
Country/Territory	Germany
City	Munich
Period	14/09/18 → 14/09/18
Internet address	https://peoplecap2018.weebly.com/

Keywords

3D registration; 3D shape morphing; 3D morphable models

Cite this

@conference{33149fbd00ee42c7b83c68f713d44042,

title = "Non-rigid 3D Shape Registration using an Adaptive Template",

abstract = "We present a new fully-automatic non-rigid 3D shape registration (morphing) framework comprising (1) a new 3D landmarking and pose normalisation method; (2) an adaptive shape template method to improve the convergence of registration algorithms and achieve a better final shape correspondence and (3) a new iterative registration method that combines Iterative Closest Points with Coherent Point Drift (CPD) to achieve a more stable and accurate correspondence establishment than standard CPD. We call this new morphing approach \emph{Iterative Coherent Point Drift} (ICPD). Our proposed framework is evaluated qualitatively and quantitatively on three datasets: Headspace, BU3D and a synthetic LSFM dataset, and is compared with several other methods. The proposed framework is shown to give state-of-the-art performance.",

keywords = "3D registration; 3D shape morphing; 3D morphable models",

author = "Hang Dai and Pears, {Nicholas Edwin} and Smith, {William Alfred Peter}",

year = "2018",

month = sep,

day = "14",

language = "English",

note = "ECCV 2018: PeopleCap Workshop ; Conference date: 14-09-2018 Through 14-09-2018",

url = "https://peoplecap2018.weebly.com/",

}

TY - CONF

T1 - Non-rigid 3D Shape Registration using an Adaptive Template

AU - Dai, Hang

AU - Pears, Nicholas Edwin

AU - Smith, William Alfred Peter

PY - 2018/9/14

Y1 - 2018/9/14

N2 - We present a new fully-automatic non-rigid 3D shape registration (morphing) framework comprising (1) a new 3D landmarking and pose normalisation method; (2) an adaptive shape template method to improve the convergence of registration algorithms and achieve a better final shape correspondence and (3) a new iterative registration method that combines Iterative Closest Points with Coherent Point Drift (CPD) to achieve a more stable and accurate correspondence establishment than standard CPD. We call this new morphing approach \emph{Iterative Coherent Point Drift} (ICPD). Our proposed framework is evaluated qualitatively and quantitatively on three datasets: Headspace, BU3D and a synthetic LSFM dataset, and is compared with several other methods. The proposed framework is shown to give state-of-the-art performance.

AB - We present a new fully-automatic non-rigid 3D shape registration (morphing) framework comprising (1) a new 3D landmarking and pose normalisation method; (2) an adaptive shape template method to improve the convergence of registration algorithms and achieve a better final shape correspondence and (3) a new iterative registration method that combines Iterative Closest Points with Coherent Point Drift (CPD) to achieve a more stable and accurate correspondence establishment than standard CPD. We call this new morphing approach \emph{Iterative Coherent Point Drift} (ICPD). Our proposed framework is evaluated qualitatively and quantitatively on three datasets: Headspace, BU3D and a synthetic LSFM dataset, and is compared with several other methods. The proposed framework is shown to give state-of-the-art performance.

KW - 3D registration; 3D shape morphing; 3D morphable models

M3 - Paper

T2 - ECCV 2018: PeopleCap Workshop

Y2 - 14 September 2018 through 14 September 2018

ER -